Motor dynamics encoding in the rostral zone of the cat cerebellar flocculus during vertical optokinetic eye movements

The complex spike (CS) and simple spike (SS) activities of Purkinje cells i n the rostral zone of the cerebellar flocculus were recorded in alert cats during optokinetic responses (OKR) elicited by a stimulus sequence consisti ng of a constant-speed visual pattern movement in one direction for 1 s and then in the opposite direction for 1 s. The quick-phase-free trials were s elected. Ninety-eight cells were identified as rostral zone cells by the di rection-selective CS activity that was modulated during vertical but not ho rizontal stimuli. In most of the majority population (88 cells), with an in creasing CS firing rate during upward OKR and an increasing SS rate during downward OKR, the inverse dynamics approach was successful and the time cou rse of the SS rate was reconstructed (mean coefficient of determination, 0. 70 and 0.72 during upward and downward stimuli, respectively) by a linear w eighted superposition of the eye acceleration, velocity, position, and cons tant terms, at a given time delay (mean 10 ms) from the unit response to th e eye-movement response. Standard regression coefficient (SRC) analysis rev ealed that the contribution of the velocity term (mean SRC 0.98 for upward and 0.80 for downward) to regression was dominant over acceleration (mean S RC 0.018 and 0.058) and position (-0.14 and -0.12) terms. The velocity coef ficient during upward stimuli (6.6 spikes/s per degree/s) was significantly (P<0.01) larger than that during downward stimuli (4.9 spikes/s per degree /s). In most of the minority population (10 cells), with both CS and SS fir ing rates increasing during upward OKR, the inverse dynamics approach was n ot successful. It is concluded that 1) in the cat rostral zone Purkinje cel ls, in which the preferred direction is upward for CS and downward for SS, eye velocity and acceleration information is encoded in SS firing to counte ract the viscosity and inertia forces, respectively, on the eye during vert ical OKR; 2) the eye position information encoded in SS firing is inappropr iate fur counteracting the elastic force; 3) encoding of eye velocity infor mation during upward OKR is quantitatively different from that during downw ard OKR: SS firing modulation is larger for upward than for downward OKR of the same amplitude; and 3) encoding of motor dynamics is obscure in cells in which the preferred direction is upward for both CS and SS.